On behalf of K-lab and the Munch Museum we are pleased to invite you to a workshop on «Text recognition and transcription of handwritten text» at the National Archives of Norway in Oslo on 13th – 14th November 2017. The workshop will focus on recent developments in the area, such as keyword spotting in digitalised handwritten text, combined with the experiences by those institutions who have tried out the technology. To facilitate an open discussion about this type of technology we have decided to extend the invitation to other Nordic GLAM institutions.
The Workshop will consist of presentations from GLAM-colleagues and the European research project Recognition and Enrichment of Archival Documents (READ, https://read.transkribus.eu/), and several hands-on sessions trying out Transkribus and other tools for working with digitalised handwritten archives.
K-lab is a cooperation between the National Archives of Norway, the Norwegian Mapping Authority, Arts Council Norway and the Norwegian Directorate for Cultural Heritage to experiment with digital presentation and dissemination and improving access to data from cultural institutions.
The Munch Museum is a Memorandum of Understanding partner in the Recognition and Enrichment of Archival Documents project (READ) and has tested out Transkribus on Edvard Munch’s letters. The Museum have digitalised Edvard Munch’s, mostly handwritten, letters and notes and made the material available online at http://emunch.no/
Day 1. 11.00–18.00
1) Short round of introductions. Getting to know the participants and their issues/objectives
2) Handwritten Text Recognition and more. Introduction to the READ project
Transkribus in practice. From the Middle Ages to 20th century, from Arabic to Latin, from Danish to Italian
How to reuse existing transcripts as training data for Automated Text Recognition
How to create training data from scratch
Transkribus workshop (coffee, tea and fruit available – short break if needed).
– How to set up a project in the Transkribus client
– How to use the web-based editor
– How to use Transkribus to set up crowd-sourcing projects
– How to use the API of the Transkribus platform
The workshop will include hands-on work in Transkribus using test sets already installed. Participants who would like to test on their own material should bring this along and will get help to get started.
Short, invited responses that discuss their practice today and the potential benefit of including (or switching to) Transkribus in their existing workflow
Day 2. 09.00–14.00
Demo. How to query documents with Keyword Spotting (instead of full-text search)
Hands-on session. How Transkribus supports the use of mobile phones as document scanners in archives. Please bring along – if you have – paper facsimiles of material you would like to test on.
Demo. TranskribusLearn. How volunteers and students can teach themselves to read old scripts
Short, invited response that discusses their practice today and the potential benefit of including (or switching to) TXT2IMG tool in their existing workflow
Sustaining Transkribus. Discussion and closing remarks
Meet the READ team. Time for workshop participants to discuss ideas for cooperation, participation, tools etc. with the representatives from READ; or to discuss and have another look at presented tools while talking to the experts from READ.
There is no attendance fee, but registration is mandatory and the registration deadline is November 1st. Participants should register via email to email@example.com
When you register, please include the following information:
- Your name
- Your title/occupation
- Your institution
- The email address you wish to use for your Transkribus account as accounts for training will be set up ahead of the workshop. If you already have a Transkribus account please state the email address it is registered on
- Any special dietary concerns
- And finally, to facilitate conversation and inspire future cooperation among participants we are including a little more info on the name tags than you usually see. Please include – in no more than 140 characters and preferably in keyword form – a brief description:
- If you are a regular participant; what do you wish to gain from your participation? Why are you here? What tool/s / functionality/ies are you most interested in?
- If you represent READ; what is your expertise in this setting? What can you offer to the participating institutions?